The Top 100 Consumer AI Apps: Who’s Actually Paying?

Episode Summary

Executive Summary: A16Z’s Top 100 Consumer AI Apps report shows consumer AI is maturing into a power-user market: traffic is stabilizing, but spending is highly concentrated among developers, creators, and technical users. The conversation highlights the rise of personal agents, the divergence of ChatGPT/Claude/Gemini, and the need for new business models beyond subscriptions—especially ads, transactions, and marketplaces.

Main Topics: Consumer AI is becoming a power-user market (Priority: 5/5): Traffic growth is leveling off while spend is concentrated among a small group of heavy users. The report added consumer card-spend data for the first time, revealing many high-spend products not visible in traffic rankings. The rise of personal agents (Priority: 5/5): A major new trend is consumer-facing assistants and agents that do actions for users, not just answer questions. Products like Muse, Instinct, and OpenAI’s newer assistant efforts signal a shift from prompt boxes to task execution. Spending is highly concentrated among builders and creators (Priority: 5/5): Most consumer AI revenue comes from developers, creators, and technical power users who use tools for coding, productivity, and content creation. The top 1% of payers spend exceptionally large amounts per month. Business models beyond subscriptions are still underdeveloped (Priority: 5/5): The speakers argue consumer AI cannot rely mainly on subscriptions forever. Ads, transaction fees, and marketplaces will be needed as costs fall and mainstream adoption broadens. ChatGPT, Claude, and Gemini are diverging (Priority: 4/5): ChatGPT remains the dominant consumer product, Claude is overperforming in paid subscribers relative to its install base, and Gemini is lagging in consumer monetization despite Google’s distribution advantage. Creative tools and specialized vertical products are thriving (Priority: 4/5): Audio, image, video, and design tools show that focused products can outperform generalist labs in specific use cases. Examples include Suno, ElevenLabs, Midjourney, and Figma-like workflows. White space remains in social, dating, shopping, entertainment, and other networked categories (Priority: 4/5): Most AI products today help users save time; the next major wave may come from products that help people spend time, create socially, or participate in multiplayer networks and marketplaces.

Key Arguments: Consumer AI traffic is settling, but spending data reveals a much more dynamic market than traffic alone shows. The consumer AI economy is a power-law game: a small number of users drive most revenue. Personal agents are the biggest new trend because they move AI from conversation to action. Mainstream adoption is constrained by trust, privacy, and safety concerns when AI accesses email, credit cards, and personal context. Subscriptions are an unnatural primary business model for consumer internet products; ads and transaction fees will likely matter more as AI matures. OpenAI’s ad run rate suggests AI-native ad products can scale quickly because chat systems have rich user context and commercial intent. Labs excel at baseline capabilities, but startups can win by packaging workflows, context, and user experience into richer products. Many large consumer categories remain open because they require network effects, marketplaces, or “spend time” experiences rather than pure productivity. Specialized tools can outperform generalist models in audio, design, and other modalities because they offer deeper workflows and stronger taste or domain focus.

Data Points: Americans reporting AI use: about half - Broad consumer awareness/adoption of AI in the U.S. Americans paying for AI: around 4.5% of U.S. consumers - Share of consumers paying for an AI subscription/product via the panel data Alternative estimate of AI payers: as low as 2% to 2.5% - Other sources cited for consumer AI subscription penetration Top 1% monthly spend: $903 per month - Top 1% of paying users’ personal AI spend on credit cards Median monthly spend among payers: $25 per month - Median spend in the consumer AI spend panel Revenue concentration in top 10% of payers: more than half of revenue - Top 10% of paying users drive the majority of consumer AI revenue Revenue share of top 1% of payers: 20% of revenue - Among paying users, the top 1% account for a large share of spend Revenue share of bottom 50% of payers: 16% of revenue - Lower-spending half of payers contributes a comparatively small share Consumer AI products ranked for spend: 50 products - Products included in the report’s spend ranking Spend-ranked products missing from traffic lists: 29 of 50 - Many high-spend products were not top traffic products New products on web + mobile lists: 11 - Fewest newcomers in prior editions, suggesting market stabilization OpenAI ad annual run rate: $1 billion - Annualized advertising revenue run rate cited for OpenAI OpenAI weekly active users: 1.2 billion - User density supporting ad monetization ChatGPT traffic vs Claude: ~6x ahead on the web - Relative consumer usage gap ChatGPT traffic vs Gemini: ~2x ahead on the web - Relative consumer usage gap Claude paid subscribers vs Gemini: Claude has passed Gemini - Claude’s consumer paid subscriber count now exceeds Gemini’s in the U.S. panel Claude max-plan mix: ~7.5% of subscribers on $100+ plan - Claude has a larger share of high-tier subscribers than ChatGPT or Gemini ChatGPT/Gemini high-tier mix: ~1% on $100+ plan - Comparison point for high-tier subscription usage Muse downloads: ~5 million in ~22 days (U.S. and Canada) - Recent consumer launch traction Threads downloads for comparison: ~16 million in ~22 days (U.S. and Canada) - Benchmark Meta launch used for comparison Muse early active users: ~250,000 active users in first 12 days - Early adoption cited for Muse Instinct growth: 100,000 users, growing 10% day over day - Assistant product momentum Instinct card-linking: 40% in first three weeks - Users connecting a credit card early in product use Instinct early spend: over $1,000 on average in first month - High willingness to pay among early users OpenEvidence physician density: 50% to 60% of U.S. physicians - Example of a vertical AI product with strong audience density

Pivotal Quotes: "the most interesting big new trend is in personal agents" — Olivia Moore: Opening framing of the conversation on where consumer AI is headed "The top 1% user is spending $903 per month personally on their personal credit cards on AI." — Olivia Moore: Discussion of consumer spend concentration among power users "we have transcended the need for everyone to buy a subscription to AI" — Olivia Moore: Argument that consumer AI must expand beyond subscriptions into ads, transactions, and other models

Implications: Consumer AI is shifting from novelty to durable products for power users, but mass-market adoption needs trust, better UX, and new monetization. Startups can still win by owning context, workflow, and network effects, especially in underexplored categories like shopping, dating, and entertainment.